ADOLESCENT DEVIANCE: WHY STUDENT ROLE PERFORMANCE MATTERS
A Thesis by
Sarah J. Victory
B.A., Wichita State University, 2004
Advised by Dr. David Wright
Submitted to the College of Liberal Arts and Sciences
and the faculty of the Graduate School of
Wichita State University in partial fulfillment of
the requirements for the degree of
Master of Arts
December 2005
ADOLESCENT DEVIANCE: WHY STUDENT ROLE PERFORMANCE MATTERS
I have examined the final copy of this Thesis for form and
content and recommend that it be accepted in partial fulfillment
of the requirement for the degree of Master of Arts with a major
in Sociology.
________________________________
Dr. David W. Wright, Committee Chair
We have read this Thesis
and recommend its acceptance:
________________________________
Dr. Jodie Hertzog, Committee Member
________________________________
Dr. Brian Withrow, Committee Member
ii
DEDICATION
To My “Honey, Honey”, Dominic Alexander Victory
“The only difference in doing something and not doing something
is doing it.” (Dr. David Wright)
iii
ACKNOWLEDGMENTS
First and foremost, I thank God for His endless mercy,
grace, and strength; for His foolishness is wiser than my wisdom.
Thanks to Dr. Wright for being an outstanding advisor and
teacher; no other individual could have been so dedicated and
helpful through this journey.
Thanks to my committee members,
Dr. Jodie Hertzog and Dr. Brian Withrow for their input and
advice to improve this thesis.
I would also like to thank Dr.
Ron Matson and Linda Matson for their never-ending support over
the last 5 ½ years. Thanks to my Sociology cohort for all of the
laughs and for making the lab a fun place to live.
Thanks to my
good friend, Eric London, for always having an answer to my
questions.
Thanks to my extended family and church family for
their continuing prayers and support.
Last, but not least, I
would like to thank my family—Mom, Dad, and Steph—for loving me,
supporting me, convincing me to never give up, and making me
believe in myself.
iv
ABSTRACT
The focus of this research project is on the relationship of
student role performance and deviant behavior in adolescents.
I
use the Education Longitudinal Study of 2002 to analyze the
relationship of the given alternative model across student role
performance factors, school factors, and family factors.
The
first two hypotheses deal with the student role performance
level, stating that males are more likely to be deviant than
females, and students who are held back a grade are more likely
to be deviant.
The next two hypotheses deal with the school
level, stating that students in schools with increased rules will
more likely be deviant and higher teacher-student ratios will
increase deviance.
The final two hypotheses deal with the family
level and state that as family SES increases, deviance decreases,
and students in two-parent families will have less deviance than
students in single-parent families.
The results of the analyses
revealed that the first set of hypotheses was supported. In the
second set of hypotheses, the first hypothesis was supported, but
the second hypothesis was not supported.
The final set of
hypotheses was supported. It was concluded that student role
performance had a significant effect on deviance.
v
TABLE OF CONTENTS
Chapter
Page
1.
Introduction................................................1
2.
Literature Review...........................................2
2.1 Deviance...............................................2
2.2 Student Role Performance (SRP).........................2
2.3 School.................................................5
2.4 Family.................................................9
2.5 Alternative Model.....................................13
2.5.1 Hypotheses.....................................14
3.
Data & Methodology.........................................15
3.1 Data..................................................15
3.2 Variables.............................................15
3.2.1 Dependent Variable.............................15
3.2.2 Independent Variables..........................16
3.2.2.1 SRP...................................16
3.2.2.2 School................................19
3.2.2.3 Family................................23
3.3 Methodology...........................................27
4.
Results....................................................28
4.1 Univariate & Bivariate Analyses.......................28
4.2 Multivariate Analysis.................................34
5.
Discussion/Conclusion......................................35
6.
Bibliography...............................................39
7.
Appendices.................................................44
vi
LIST OF TABLES
Table
Page
1.
Univariate & Bivariate Analysis............................45
2.
Univariate & Bivariate Analysis Across Family Structure....47
3.
Multivariate Analysis......................................49
vii
1. Introduction
Educational success is a determinant of an individual’s
likelihood to attend college and obtain a good job.
The
relationship between educational success and career achievement
is demonstrated by the national policy No Child Left Behind (US
Government, 2001).
This policy measures the performance of
schools and teachers based on national standardized test scores
of students.
However, certain factors can cause a regression in
academic success, such as deviance which can negatively affect
student role performance.
The future success of students will be
determined by understanding and acknowledging the factors
responsible for hindering student role performance.
Previous literature addresses different factors that can
affect the likelihood of an adolescent engaging in deviant
behavior such as sex, educational commitment, attendance
patterns, and handicaps.
School demographics and rules, parental
involvement, class sizes, and school environments have also been
noted as factors that relate to deviance in adolescents.
Past
studies have indicated that socioeconomic status, family
demographics, family resources, communication, and family rules
can also be factors in relation to deviance.
All of these
factors listed can be put into three main categories: 1) student
role performance (SRP), 2) school factors, and 3) family factors.
This study uses the 2002 Education Longitudinal Survey (ELS) to
1
examine the influence that each level has on adolescent deviance.
2. Literature Review
2.1 Deviance
Previous studies conducted on deviance have generated
several theories.
One theory is the Akers Social Learning Theory
where deviance is defined as violent, destructive, and illegal
behavior.
This theory states that the likelihood for adolescents
to engage in deviant behavior is affected by their attitudes
regarding deviance, their imitations of people around them, and
the consequences (positive and negative) that result from deviant
behaviors (Ardelt & Day, 2002).
The Control Theory suggests
that deviant behavior can be minimized by strong relationships
between adolescents and society (Downs, Robertson, & Harrison,
1997).
Labeling Theory suggests that if adolescents are labeled
as “deviant”, they are likely to feel stigmatized and become even
more engaged in deviant behavior (Downs, Robertson, & Harrison,
1997).
2.2 Student Role Performance
Student role performance (SRP) refers to the indicators by
which institutional agents assess students’ abilities to meet
given expectations.
SRP is affected by these agents. Components
of SRP include students’ sex, race, language, time spent on
homework, school attendance, and handicap factors.
An adolescent’s sex can have an impact on their likelihood
2
of becoming involved in deviant behavior. One cause of this is
due to the fact that adolescent males and females usually
contrast in terms of maturity and social interaction (Crosnoe,
Erickson, & Dornbusch, 2002). Studies have shown that adolescent
girls are less deviant, have fewer deviant friends, and are less
likely affected by their friends’ negative behaviors than boys
(Crosnoe, Erickson, & Dornbusch, 2002). Both boys and girls who
have low self-control and lack social behaviors growing up are
likely to engage in deviant behavior as adolescents (Mason &
Windle, 2002).
Adolescents’ race/ethnicity did not result as having any
significant impact upon acts of deviant behavior (Dornbusch,
Erickson, Laird, & Wong, 2001). However, race and ethnicity can
influence academic performance, which can eventually lead to acts
of deviance. African Americans and Caucasians tend to be less
committed to their education than Hispanics (Johnson, Crosnoe, &
Elder, 2001).
According to Epstein (1972), African Americans,
Hispanics, and Native Americans have low educational attainment
as a result of being minorities in society.
Adolescents vary in their levels of educational commitment,
such as spending time doing homework, keeping good grades,
desiring to pursue college, etc. One study found that the more
time adolescents spend on homework and keeping good grades, the
less likely they are to have friendships with others involved in
3
deviant acts (Erickson, Crosnoe, & Dornbusch, 2000). A high level
of educational commitment usually signifies the absence of
delinquency and the tendency to obey given rules (Erickson,
Crosnoe, & Dornbusch, 2000).
Attendance patterns at school and the possibility of grade
retention affect the likelihood of adolescent deviance.
Adolescents who do not establish good attendance patterns usually
become truant pupils (Bonikowske, 1987).This puts them at a
higher risk of becoming involved in such deviant behaviors as
gang activity, smoking, and drinking (Garry, 1996). Truancy also
causes students to get behind in homework assignments, and the
further behind they get, the more likely they are to eventually
drop out of school (Bonikowske, 1987). If a child experiences
grade retention or begins school at a later age, then they are at
a higher risk for future behavior problems (Byrd, Weitzman, &
Auinger, 1997).
The reason for this higher risk level is due to
the fact that adolescents who are older than the rest of their
peers may not feel as appropriate in socializing with them.
Therefore, they are likely to engage in other devious behaviors.
(Byrd, Weitzman, & Auinger, 1997.) They may also feel like they
do not meet up to their peers’ levels of intelligence, which can
cause them to fall behind in homework and completely disengage
themselves from school. (Byrd, Weitzman, & Auinger, 1997).
Handicaps and disabilities can be determining factors in
4
whether or not an adolescent is likely to engage in deviant
behavior.
This can include feelings of inadequacy, physical
disabilities, and mental disabilities (Dekker & Koot, 2003).
Physical handicaps, such as speech or hearing impairments, can
cause barriers in communication and making friends for
adolescents (Redmond & Rice, 1998).
These barriers can result in
teachers retaining a child, which can ultimately result in
deviance, as mentioned earlier (Byrd, Weitzman, & Auinger, 1997).
Individuals with disabilities tend to not be accepted by others,
especially by their peers in school.
This feeling of low
acceptance can be the cause of loneliness, truancy, quitting
school, and juvenile delinquency (Cook & Semmel, 1999).
2.3 School
The composition of schools has been shown in some studies to
have an effect on deviance. School demographics, environments,
rules, problems, and programs should all be taken into account
concerning adolescent deviance. Factors such as public versus
private schools, strict rules, classroom environment, studentteacher ratios, teacher-student relationships, low income, school
violence, and physical and sexual abuse, can determine whether or
not a student becomes involved in deviant behavior.
As more parents become leery of school safety and the lack
of discipline enforcement in public school systems, more students
are being enrolled in private schools (Schaller, 1979).
5
Differences between public and private schools can play a factor
in whether a child is likely to be deviant or not.
One
difference is the level of parental involvement, which plays a
significant role in students’ achievement (Rothstein, 2000).
More parental involvement is usually found in private schools, as
some private schools mandate parents to sign an agreement stating
they will be involved in a certain number of their child(ren)’s
activities (Rothstein, 2000).
Rules and regulations vary from district to district, and
rules also vary between public and private schools.
have much more rigid rules than others.
Some schools
It is important, as
with any structured organization, to have rules and guidelines
that are to be followed (Sampson and Raudenbush, 1999).
However,
it has been shown that some schools can actually be too strict,
causing a negative reaction from the students.
Intense rule
enforcement can make students feel uncomfortable in their
environment (Sugai, Sprague, Horner, & Walker, 2000).
For
instance, the use of metal detectors, security officers,
mandatory hall passes, and uniform policies can increase the
structure of a school, but adolescent students tend to react
negatively to such rules (Sugai, Sprague, Horner, & Walker,
2000).
These devices, along with locker and bag searches, can
increase fear and anxiety in students, and they do not address
the main reason or motivation behind students carrying weapons in
6
the first place (Juvonen, 2001).
Having uniformed officers can
create a feeling of apprehension among students, leading to more
acts of misbehavior (Juvonen, 2001).
Another common policy among
schools is the “zero-tolerance” policy.
Zero-tolerance policies
implement punishment after any single act which violates
established rules, and those punishments are usually either
suspension or expulsion (Juvonen, 2001).
This type of policy may
reduce problems inside the school building, but it also decreases
the liklihood of academic success for the student, as suspension
and/or expulsion gives time and opportunities for delinquency
involvement and dropping out of school permanently (Juvonen,
2001).
Student-to-teacher ratios and class sizes can affect a
student’s academic performance and behavior.
There is a high
amount of diversity in classrooms, and a reasonable student-toteacher ratio needs to be maintained for a productive classroom
environment (Farrelly, 2001).
However, most schools lack the
needed funding to balance those ratios appropriately (Farrelly,
2001).
When an appropriate ratio is established, teachers can
more easily build personalized relationships with their students,
which will increase students’ efforts in learning (Cresswell &
Rasmussen, 1996).
Research supports that student-to-teacher
ratios have a fair amount of impact on student achievement
(Murray, 2000).
7
The school environment as a whole influences students’
academic success.
Factors such as violence, absenteeism,
vandalism, and disrespect can create poor learning environments.
School violence can cause students to form negative feelings
towards attending school, especially if they do not feel safe in
their environment.
Almost 25% of students in public school
systems report victimization of violence acts while being at
school (Bennett-Johnson, 2004).
School violence can be
associated with gang activity, flagrant disregard of studies, and
void of interest in class (Bennett-Johnson, 2004).
Some schools
implement programs to help reduce violence, such as having crisis
response teams, on-site counselors, and designated staff members
who are responsible for reporting incidents to authorities and
notifying parents (SVRC, 2002).
Community support and
involvement from parents, staff, and students can also aid in
minimizing violence in schools (Bennett-Johnson, 2004).
Vandalism in schools is a common practice by deviant students,
especially those involved in gangs (Pappalardo, 2002).
Schools
used for outside community activities have high susceptibility of
vandalism (Pappalardo, 2002).
Vandalism prevention provides
students with a safer environment and can be implemented by using
hidden cameras, eliminating school lockers, monitoring parking
lots, and having only one accessible entrance to the school
building (Pappalardo, 2002).
8
2.4 Family
There are a number of family-related categories that affect
adolescent deviance, including socioeconomic status (SES), family
demographics, family resources, family communications, and family
rules. Each of these categories has its own factors which
contribute to a child’s likelihood of being involved in deviant
activities.
Family demographics consist of such things as SES,
family size, household language, employment, parents’ education,
single parenting, and marital status.
SES is an extremely important factor in the academic
performance of students (Wenglinsky, 1997).
Low family income
can result in a child becoming involved in deviant activities
such as stealing or selling drugs (Hofferth, Smith, McLoyd &
Finkeistein, 2000). Obviously, the higher the SES of a family,
the better the child will be capable of performing, academically.
This is because a child from a high SES household will have
access to a high number of various resources, such as having a
computer, having a dictionary, having access to a library, and
having the ability to afford extracurricular activities
(Hofferth, Smith, McLoyd & Finkeistein, 2000).
As for family size, the smaller a family, the more likely
the student is to perform well in school (Greenberg, 1985).
Families with fewer children can provide necessary study
resources and increase the likelihood of satisfactory educational
9
performance (Downey, 1995). According to Douglas Downey (1995),
as families grow in size, the resources available for children
are reduced, which can cause poor academic performance.
A large
family size is a good predictor that those children are likely to
obtain a very low education for reasons such as less one-on-one
time with parents, and less help on homework from parents (Riala,
Isohanni, Jokelainen, & Jones, 2003).
Children may become disengaged in school if the language
spoken at home is not the same as the language used by their
teachers.
This has become a rising issue, seeing that close to
three million children are involved in English as a Second
Language (Shore, 2001). A California study found that 62% of the
lowest test score results belong to kids who are learning English
as a second language, showing that language barriers can have a
significant impact upon student performance (Wells, 2001).
Greenberg (1985) points out that the higher the education of
the father in the household, the more likely the child is to
attain an even higher educational level.
A mother’s educational
level has a weaker influence on educational aspirations of males
than a father’s education (Sewell & Shah, 1968).
If a mother is
the only parental component in a household, then her education
level is a strong predictor of the child’s academic achievement.
(Matherne & Thomas, 2001). Sewell & Shah (1968) point out that
both parents’ educational attainment effect the educational
10
desires of females.
Children of single-parent households engage in deviant
behavior much more often than those children from a two-parent,
married household (Downey, 1994). Typically, an adolescent from a
single-parent household has less supervision than those from a
two-parent household due to the amount of time spent working by
the parent (Garasky, 1995).
As a result, this lack of
supervision prevents necessary emotional bonding and attachment
for healthy family relationships within the household (Matherne &
Thomas, 2001).
The family resources available to students can affect their
overall performance and well-being.
Those resources include
items such as a dictionary, encyclopedia, computer, books, access
to a library, calculator, washer, dryer, microwave, and study
space.
It is likely that low-income households will be able to
provide fewer resources for the children in the family (Hofferth,
Smith, McLoyd & Finkeistein, 2000). Other family resources which
may deter children from deviant behavior include access to
religious or community activities.
Adolescents who are involved
in religious or community activities are not likely to involve
themselves in deviant acts (Erickson, Crosnoe, & Dornbusch,
2000).
Communication plays an important role in establishing good
relationships between parents and children.
11
Parents who
communicate regularly with the school and teachers regarding
their child’s performance are more likely to have a productive
and open relationship with that child (Chyung, Darling, &
Caldwell, 1998).
Parents can also encourage their children by
discussing future educational plans with them and by being
involved with their homework, studies, and school programs
(Rothstein, 2000).
Lastly, family rules can influence the way a child chooses
to behave outside of the home.
Such rules as getting homework
checked by parents, limiting television/video game time, and
requiring chores to be done are household rules that some
families implement.
Some parents may also establish a curfew for
the child, require homework to be completed before other leisure
activities, and insist that good grades be maintained (Chyung,
Darling, & Caldwell, 1998).
Those children who feel they have an
obligation to follow a set of rules enforced by their parents are
unlikely to establish friendships with others who are deviant,
which decreases their susceptibility in becoming involved in
deviant behavior themselves (Chyung, Darling, & Caldwell, 1998).
12
2.5 Alternative Model
Schools
Family
SRP
(adapted from Wright, 2005)
Deviance
In this study, deviance is based on negative behavior
displayed by adolescents in school.
This negative behavior means
the student got in trouble for some act that may have resulted in
disciplinary action by school administration.
Student Role Performance (SRP) has a direct effect on
deviance. SRP is defined by the indicators used by institutional
agents to gauge a student’s ability to meet expectations that
have been set.
SRP is influenced by factors such as sex, race,
time spent on homework, school attendance, involvement in
religious and community activities, and handicaps.
For example,
lack of study time and class preparation can result in low SRP,
which in turn, increases the likelihood of a student to become
involved in deviant behavior.
The school is the structural segment of the alternative
model, as it is the institution that defines deviance.
Factors
such as private versus public schools, strict rules, teaching
styles and student expectations, classroom environments, student-
13
teacher ratios, teacher-student relationships, low income, school
violence and abuse directly impact SRP.
Thus, the school affects
not only the definition of deviance, but also the SRP of the
student.
The third segment of this alternative model is the family.
Family factors such as socioeconomic status (SES), family
demographics (family size, class, and parental education)
resources, communication, and rules directly affect SRP.
If the
parental education and SES are high, it is likely that education
will be emphasized in that household.
Family resources
(dictionaries, computers, calculators, study space, etc.) are
most likely going to be available in households with a higher
class and SES, reducing the likelihood for the student to engage
in deviant behavior.
2.5.1 Hypotheses
1a. Males are more likely to be deviant than females, net of
other factors.
1b. Net of other factors, students who are held back a grade
are more likely to be deviant.
2a. Net of other factors, students in schools with increased
rules will more likely be deviant.
2b. Net of other factors, higher teacher-students ratios
will increase deviance.
3a. Net of other factors, as family SES rises, deviance
decreases.
3b. Net of other factors, students in two-parent families
will have less deviance than students in single-parent
families.
14
3. Data & Methodology
3.1. Data
Data for this research is taken from the Education
Longitudinal Study (ELS) of 2002 sponsored by the National Center
for Education Statistics (NCES) of the United States Department
of Education.
The sample for the ELS was taken from a national
population of 10th graders who were studied throughout high
school.
There were 752 schools that participated in the study,
with approximately 26 students from each school, giving a final
sample size of 17,591 participants.
Sample restrictions were
placed to select out only those participants who had completed
questionnaires for each model segment of the study, resulting in
a final sample size of 11,046 participants for this research.
The use of standard weights in the ELS allows for an
accurate display of the entire population in the study.
Since
weights usually impose a high number of sample sizes that distort
the results by way of collapsing the standard error, a relative
weight must be used.
The standard weight is used to attain the
sample size, but the relative weight is used in order to discuss
bivariate and multivariate analyses without producing biased
results.
3.2. Variables
3.2.1. Dependent Variable
The dependent variable for this research is deviance.
The
variable chosen to represent deviance is the variable labeled
15
‘how many times student got in trouble’ in the ELS. It is a
binary variable coded as 0 for ‘not deviant’ and 1 for ‘deviant’.
3.2.2. Independent Variables
3.2.2.1. Student Role Performance (SRP)
The first model segment for this research project addresses
student role performance (SRP) variables including: sex, race,
language, grade retention, remedial courses, participation in
extra-curricular activities, participation in college preparation
programs, handicaps, class preparation, homework, hours worked
each week, and hours spent watching television/DVD.
The sex variable is a binary variable that identifies the
sex of the participant.
and 1 for ‘female’.
This variable is coded as 0 for ‘male’
It is expected that males will be shown to
have a higher level of deviance than females.
The race/ethnicity variable is a nominal variable, and it
measures the race/ethnicity of the respondent.
For this
variable, 1 is ‘White, non-Hispanic’, 2 is ‘Black, non-Hispanic’,
3 is ‘Hispanic’, 4 is ‘Asian, non-Hispanic’, and 5 is ‘other,
non-Hispanic’.
A binary variable was created from the
race/ethnicity variable to measure the respondent’s minority
status. The minority variable is coded as 0 for ‘not a minority’
and 1 for ‘minority’.
It is suspected that these variables can
influence the SRP of individuals, and the higher levels of
deviance will be seen among minorities.
16
The non-English variable is a binary variable that
identifies whether or not English is the respondent’s native
language.
It is coded as 0 for ‘those whose native language is
English’ and 1 for ‘those whose native language is not English’.
The language spoken by each student can possibly affect their
SRP, and it is suspected that those who do not speak English will
have higher rates of deviance due to language barriers.
To determine whether or not the respondent had ever been
held back a grade, a binary variable is created where 0 is for
‘those who have never been held back’ and 1 is for ‘those who
have been held back at least one grade’.
This variable is
expected to show an increase of deviance in those students who
have been retained a grade level.
The remedial math and remedial reading variables are binary
variables that determine whether or not the respondent ever took
a remedial math or reading course, coded as 0 for ‘no’ and 1 for
‘yes’.
Then, an index of both variables is created to form a
binary variable in order to determine if the respondent had any
(either math of reading) remedial courses with the same coding as
above, 0 for ‘no’ and 1 for ‘yes’.
The presence of remedial
courses can be an identifying factor of those students who are
more likely to be deviant, and it is suspected that a raise in
deviance will be apparent for these students.
Several variables in the ELS refer to various student
17
impairments including: learning disabilities, speech/language
impairments, mental retardation, emotional disturbance, hearing
impairments, orthopedic impairments, visual impairments, and
other disabilities.
These variables are combined to create an
index for a handicap variable.
The handicap variable is binary,
coded as 0 for ‘no handicap’ and 1 for ‘handicap’.
It is
presumed that this variable will have an effect on deviance, as
students with disabilities tend to fall away from meeting normal
expectations set by the school, which in turn will cause them to
be deviant.
The homework variable consists of several ELS variables that
measure the number of hours per week spent on homework for
various subjects.
The constructed variables ranges from 0 to 21
or more hours spent per week on homework.
It is suspected that
more time spent on homework will increase the student’s SRP
level, which will show significance in the student being less
likely to engage in deviant behavior.
The class preparedness variable is created by using several
variables that represent class preparation to make a scale.
The
variables included in the scale were how often the student goes
to class without a pencil or paper, how often the student goes to
class without a book, and how often the student goes to class
without having homework completed.
scale is .770.
The Cronbach alpha for this
The variable is coded where 1 means ‘never not
18
prepared’, 2 means ‘seldom not prepared’, 3 means ‘often not
prepared’, and 4 means ‘usually not prepared’.
Therefore, a
higher number indicates that the student is not prepared for
class.
It is expected that the students who are usually prepared
for class will not have high levels of deviance.
The extracurricular activities variable measures the number
of hours per week spent on extracurricular activities such as
sports, music, or other community activities.
The variable
ranges from 0 to 21 or more hours spent per week on
extracurricular activities.
It is suspected that more time spent
on extracurricular activities will decrease the likelihood of
deviant behavior for the respondent.
The ELS contains variables relating to how many hours the
respondent works during the week and during the weekend.
These
variables are used to create an index to measure how many hours
the respondent works in an entire week, and then the variable is
recoded to measure whether or not the respondent actually had a
paying job.
This variable is binary, with 0 as ‘not having a
paid job’ and 1 as ‘having a paid job’.
It is presumed that the
respondents with paid jobs will be less likely to engage in
deviance.
3.2.2.2. School Level
The second model segment for this research project addresses
school-level variables including: private school, school size,
19
class size, percent of students in free lunch program, percent of
students in college preparation programs, percent of students in
ESL, school problems, negative school environment, and school
rules.
The private school variable is a binary variable that shows
whether or not the respondent attended a private school.
variable is coded as 0 for ‘no’ and 1 for ‘yes’.
The
It is suspected
that the results from this variable will show that there is a
lower amount of deviance in private schools than in public
schools.
The variable for school size is an interval level variable
ranging from 200 to 2500 students.
This variable will give an
idea of how many students were in attendance at the respondent’s
school, with the expectation that more deviance will occur in
larger schools due to larger class sizes and less attention per
student.
A variable for the class size of the 10th grade is
interval level, ranging from 50 to 700 students.
It is presumed
that smaller class sizes will result in low levels of deviance.
An interval level variable is created to show the percent of
students on a free lunch program at the respondent’s school.
The
amount of students receiving free lunch can represent a lower
income status among those students.
It is suspected that schools
with higher percentages of students on free lunch programs will
have higher rates of deviance.
20
The variable for the percent of students in a college prep
program is an interval level variable.
This variable is expected
to show that schools with more students who are involved in
college prep programs will have lower levels of deviance.
Another interval level variable is created to show the
percentage of students in the respondent’s school who are
receiving English as a Second Language (ESL) instruction.
Those
students who do not understand English are very likely to not
understand the expectations of the school, which is expected to
result in a higher rate of deviance.
The ELS contains several variables regarding how often
certain problems arise at the school.
Those problems include:
tardiness, absenteeism, class cutting, physical conflicts,
robbery/theft, vandalism, alcohol use, illegal drug use, students
on drugs/alcohol at school, sale of drugs near the school,
possession of weapons, physical abuse of teachers, verbal abuse
of teachers, racial tension among students, student bullying,
disorder in classrooms, student disrespect for teachers, gang
activity, and cultextermist group activities.
These variables
are used to create a scale for an interval level variable that
measures school problems.
The variable ranges from 1 to 5, and
it is coded as 1 for ‘never happens’, 2 for ‘happens on
occasion’, 3 for ‘happens at least monthly’, 4 for ‘happens at
least weekly’, and 5 for ‘happens daily’.
21
The Cronbach alpha for
the scale is .86.
School problems indicate deviant misbehavior,
therefore, it is expected that high numbers for this variable
will result in high levels of deviance.
Several variables used in the ELS describe a negative school
environment that hinders learning.
Those variables include:
poor conditions of the building, poor heating/air/light, poor
science labs, poor fine arts facilities, lack of space, poor
library, lack of texts/supplies, too few computers, lack of
multi-media, lack of discipline/safety, and poor vocational/tech
equipment and facilities.
A scale was used to create a new,
interval level variable ranging from 1 to 4 based on the amount
of negative conditions.
The Cronbach alpha is .91.
It is coded
as 1 for ‘not at all’, 2 for ‘very little’, 3 for ‘some extent’,
and 4 for ‘a lot’.
Consequently, a higher number represents a
really poor school environment.
Since poor school environments
are responsible for learning hindrances, it is suspected that
those schools with negative environments will have an increase in
deviance.
The final variable in the school segment model is school
rules.
This variable was created from an index of ELS variables
regarding certain rules pertaining to the school.
Those
variables include: control access to buildings during school
hours, control access to grounds during school hours, require
students to pass through metal detectors, random metal detector
22
checks on students, close campus for students during lunch,
random dog sniffs to check for drugs, random sweeps for
contraband, require drug testing for any students, require
student to wear uniforms, enforce strict dress code, require
clear book bags/ban book bags, require students to wear
badges/picture ID, require faculty/staff to wear badges/picture
ID, and use security cameras to monitor school.
The created
variable ranges from 0 to 12, with 0 as a low number of school
rules and 12 as a high number of school rules.
It is suspected
that the lower number of school rules will show a higher level of
deviant behavior.
3.2.2.3. Family Level
The final model segment for this research project addresses
family-level variables including: family structure, rural/urban
residency, number of siblings, years in neighborhood, parents’
education, Socio-economic status, parental involvement with
student, parental advice to student, parental involvement with
school, number of parental rules in the household, number of
siblings who have dropped out of school, family resources,
parent-student communication, parent help with homework, and
family meal times.
The family structure variable was used to create two binary
variables for two-parent families and single-parent families.
The two-parent family variable is coded as 1 for ‘two-parent’ and
23
0 for anything other than two-parent.
The single-parent family
variable is coded as 1 for ‘single-parent’ and 0 for anything
other than single-parent.
It is suspected that single-parent
families will have a greater amount of deviant behavior present.
The urban variable is a binary variable that measures
whether the student is in a rural or urban area.
0 for ‘rural’ and 1 for ‘urban’.
It is coded as
It is presumed that families in
urban areas will have higher levels of deviance than those in
rural areas.
The siblings variable is an interval level variable that
will show the number of siblings of the respondent.
variable ranges from 0 to 6.
This
It is expected that the more
siblings a student has, the more likely they will be to engage in
deviant behavior.
An interval level variable is used to show the number of
years the respondent has lived in a particular neighborhood.
This variable ranges from 0 to 50. It is presumed that those who
remain stationary for longer periods of time will tend to have
lower levels of deviance.
A binary variable is used to show whether or not a
respondent’s parent(s) have a college degree.
coded as 0 for ‘no’ and 1 for ‘yes’.
This variable is
It is expected that
respondents whose parents have college degrees will not be likely
to engage in deviant behavior.
24
An interval level variable is created using a scale of
several variables in the ELS to show involvement between the
parent(s) and student.
Those variables include: attended school
activities with 10th grader, worked on homework/school projects
with 10 grader, attended concerts/plays/movies with 10th grader,
attended sports events outside school with 10th grader, attended
religious services with 10th grader, attended family social
functions with 10th grader, took day trips/vacations with 10th
grader, worked on hobby/played sports with 10th grader, went
shopping with 10th grader, went to restaurants with 10th grader,
spent time talking with 10th grader, and did something else fun
with 10th grader.
The Cronbach alpha is .80, and the variable
ranges from 1 to 4.
It is coded as 1 for ‘never’, 2 for
‘rarely’, 3 for ‘sometimes’, and 4 for ‘frequently’.
A lower
level of deviance will be seen among those who have a high level
of parent-student involvement.
The variable for parental advice is an interval level
variable created from a scale of several variables in the ELS
regarding advice provided about: selecting courses or programs,
plans for college entrance exams, applying to college/school
after high school, jobs to apply for after high school,
information about community/national/world events, and things
troubling the 10th grader.
The Cronbach alpha for this scale is
.76, and the variable ranges from 1 to 3.
25
It is coded as 1 for
‘never’, 2 for ‘sometimes’, and 3 for ‘often’.
Those who have
more advice will be less likely to be deviant.
An interval level variable is created from a scale of
variables in the ELS relating to parents’ involvement with the
school.
Those variables include: belong to parent-teacher
organization (coded as 1), attend parent-teacher organization
meetings (coded as 2), take part in parent-teacher organization
activities (coded as 3), act as a volunteer at the school (coded
as 4), and belong to other organizations with parents from school
(coded as 5). Zero represents those parents who are not involved
with the school.
The Cronbach alpha for the scale is .71.
It is
suspected that lower levels of deviance will be seen among those
whose parents are most involved with the school.
The parent rules variable is an interval level variable
ranging from 1 to 8 which represents the number of rules the
parents enforce at home.
It is suspected that those students who
have more rules enforced at home will be less likely to engage in
deviant behavior.
An interval level variable is used to show the student’s
number of siblings who have dropped out of school.
ranges from 0 to 6.
This variable
It is expected that a higher number of
siblings who have dropped out will result in the student being
more prone to deviance.
The family resource variable is an interval level variable
26
created from a scale of several variables in the ELS. Those
variables identify families’ resources including: daily
newspaper, regularly received magazine, computer, access to the
Internet, DVD player, electric dishwasher, clothes dryer, more
than 50 books, student has his/her own room, and fax machine.
ranges from 0 to 10.
It
Those who have access to several resources
will be less likely to be deviant than those who have a limited
amount of resources.
The parent helps with homework variable is a binary variable
that identifies whether or not a parent helps the student with
their homework.
It is coded as 0 for ‘does not help’ and 1 for
‘helps with homework’.
It is suspected that those students who
receive help with homework will not be likely to engage in
deviant behavior.
The meals variable is an interval level variable that states
how many meals are eaten together as a family in the student’s
household.
This variable ranges from 0 to 7.
Those who spend
more time together eating meals will be less likely to partake in
deviant activities.
3.3 Methodology
This research project uses SPSS to run univariate,
bivariate, and multivariate analyses.
The univariate analysis is
performed to determine and identify various groups within the
study sample.
The tests consist of those such as ‘frequency’ and
27
‘descriptive’ in order to obtain the mean values of both
dependent and independent variables.
A bivariate analysis is
used to show differences between two independent variables across
the dependent variable, deviance.
A t-test is used to determine
whether or not the differences between groups are statistically
significant or not.
Multivariate analysis is used to identify
which variables are significant, net of other factors.
Logistic
regression is a multivariate analysis that is used to examine the
relationship between the binary dependent variable, deviance, and
the other independent variables in relation to family structure.
This method of multivariate analysis can determine the
probability of group membership across other independent
variables.
4.
Results
4.1 Univariate and Bivariate Analyses
Table 1 provides univariate and bivariate measures for the
mean values of those students who are deviant or not deviant.
table
also
provides
the
bivariate
measures
for
performance, school level, and family level factors.
student
The
role
Data is taken
from the 2002 Education Longitudinal Survey which is a national
probability data set of 10th graders in the United States.
Forty percent of the entire sample is deviant. Among students
who are not deviant, 59% are female (59% vs. 37%). This means the
other 41% are male, which supports Hypothesis 1. Among students who
28
are not deviant, 13% are non-English speakers (13% vs. 10%).
Students who are not deviant spend an average of 11.1 hours per
week on homework, whereas students who are deviant spend only 9.0
hours per week on homework.
Students who are not deviant spend
more time in extracurricular activities than students who are
deviant (5.97 vs. 4.02).
Among students who are not deviant, 20%
are enrolled in a college prep program (20% vs. 16%).
Among
students who are deviant, 35% percent are minority (35% vs. 30%),
16% have been held back (16% vs. 11%), 13% partake in remedial
coursework (13% vs. 9%), and 14% have a handicap (14% vs. 9%).
Students who are not prepared for class are more likely to be
deviant than those who are prepared for class (2.06 vs. 1.75).
Students who are deviant spend an average of 13.93 hours per week
working (13.93 vs. 10.47) and 10.69 hours per week on tv/dvd (10.69
vs. 9.43).
Statistical significance is seen in all of the SRP
factors.
On the school level, the average class size of those students
who are not deviant is larger than the average class size of
students who are deviant (355.13 vs. 336.98).
Among students who
are not deviant, approximately 61% of the schools offer a college
prep program (60.9% vs. 59.6%). The average number shown for school
problems of students who are not deviant is 2.40 (vs. 2.38).
Among
the deviant students, there is a higher rate of those who attend a
private school (11% vs. 6%).
The variables for school environment
29
and the number of school rules did not show any statistical
significance.
On the family level, the average SES percentile of students
who are not deviant is higher than those who are deviant (51.7% vs.
48.3%). Among students who are not deviant, the average number
shown for parent involvement with the student is 3.15 (vs. 3.08),
parents advising the student is 2.25 (vs. 2.21), and parents
involved with the school is 1.88 (vs. 1.78).
The average number of
resources available to students who are not deviant is 6.51 (vs.
6.30).
Among non-deviant students, a result of 0.49 is given for
those who receive help with their homework from parents (vs. 0.42).
Students who are not deviant have an average number of 5.44 for
the days per week that they have at least one meal with their
family (vs. 5.18).
Among students who are deviant, an average of
26% are from single-parent families (vs. 21%), and they have an
average of 2.37 siblings (vs. 2.29).
The average number of
parental rules for deviant students is 6.96 (vs. 6.85), and the
average for number of siblings who have dropped out is 0.26 (vs.
0.18). Statistical significance is not seen with the rural and
urban residence variables or the years in neighborhood variable.
Table 2 provides univariate and bivariate measures for the
mean values of those students who are deviant or not deviant
according
to
family
structure.
This
table
also
provides
the
bivariate measures of student role performance, school level, and
30
family level factors for single-parent versus two-parent families.
Among students who are not deviant and live in single-parent
families, 61% are female (vs. 58%), 45% are minorities (vs. 26%),
and 15% are held back at least one grade in school (vs. 10%). An
average of 1.9 is shown for students not being prepared for class
(vs. 1.72).
They spend an average of 11.48 hours per week working
(vs. 10.20) and an average of 9.81 hours on tv/dvd (vs. 9.34).
For
students who are deviant and live in two-parent families, an
average of 11.2 hours per week is spent on homework (vs. 10.6),
5.34 hours is spent in extracurricular activities (vs. 4.07),and
20% are enrolled in a college prep program (vs. 16%).
Among non-
deviant students, the variables for non-English speaking, remedial
coursework,
and
handicaps
are
according to family structure.
not
statistically
significant
For students who are deviant and
live in single-parent families, 40% are female (vs. 35%), 47% are
minorities (vs. 30%), 22% have been held back at least one grade in
school (vs. 14%), and 18% have a handicap (vs. 12%).
An average of
2.1 is shown for students not being prepared for class (vs. 2.0).
They spend an average of 11.10 hours on tv/dvd (vs. 10.54).
Among
deviant students in two-parent families, an average of 9.2 hours
per week is spent on homework (vs. 8.3), and an average of 4.31
hours is spent in extracurricular activities (vs. 3.2).
Of these
students, 16% are enrolled in college prep programs (vs. 14%).
Among deviant students, the variables for non-English speakers,
31
remedial coursework, and hours students works per week are not
statistically significant.
As for school level factors of non-deviant students in singleparent families, the average number shown for a negative school
environment is 1.79 (vs. 1.73), and the average number of school
rules is 4.81 (vs. 4.59). Seven percent of non-deviant students in
two-parent families attend a private school (vs. 4%), and 61.7%
partake in college prep programs (vs. 58.5%). Class size and school
problems are not statistically significant according to family
structure.
Among deviant students of single-parent families, an
average of 2.40 is given for school problems (vs. 2.37), an average
of 1.78 is given for negative school environment (vs. 1.71), and an
average of 4.89 is given for number of school rules (vs. 4.65).
For deviant students of two-parent families, 12% attend a private
school (vs. 7%), and 61% partake in college prep programs (vs.
55.6%).
Class size for deviant students is not statistically
significant according to family structure.
For family level factors of non-deviant students in singleparent families, 37% live in an urban residence (vs. 25%).
They
have an average of 2.43 siblings (vs. 2.26), and the average number
of siblings who have dropped out is 0.27 (vs. 0.16).
Among non-
deviant students in two-parent families, 21% live in a rural
residence (vs. 18%), and they have lived in their neighborhood for
an average of 11.11 years (vs. 9.94).
32
The average SES percentile
for these students is 54.8% (vs. 40.4%).
The average number given
for parents being involved with the student is 3.17 (vs. 3.07), and
the average number given for parents being involved with the school
is 1.97 (vs. 1.57).
The average number of family resources
available to the student is 6.75 (vs. 5.65).
An average of 0.51 is
shown for those who receive help with homework from their parents
(vs. 0.42), and they have an average of 5.51 for days a week that
the student has at least one meal with their family (vs. 5.18). For
non-deviant students, the variables for parents advising students
and number of parental rules are not statistically significant.
For deviant students in single-parent families, 33% are from an
urban residence (vs. 27%), and they have an average of 0.32
siblings who dropped out of school (vs. 0.23). Among deviant
students from a two-parent family, the average SES percentile is
51.5%
(vs.
39.2%).
The
average
number
shown
for
parental
involvement with these students is 3.11 (vs. 3.00), parental advice
to the student is 2.23 (vs. 2.15), parental involvement with the
school is 1.89 (vs. 1.46), and the number of parental rules is 6.99
(vs. 6.87).
The average number of family resources available to
deviant students in two-parent families is 6.53 (vs. 5.66).
The
average number shown for parents helping the student with homework
is 0.43 (vs. 0.38), and they have an average of 5.29 for the number
of a days a week that they have at least one meal with their family
(vs. 4.88).
Statistical significance is not found for variables
33
regarding rural residence, number of siblings, or number of years
lived in the neighborhood.
4.2 Multivariate Analysis
Table 3 provides multivariate measures for single-parent and
two-parent households using logistic regression.
This analysis
identifies which of the independent variables are identified with
deviance across family structure.
As expected, being in a single-parent family increases odds of
being deviant by 1.14 compared to that of two-parent family.
Table
3 shows that there is a difference of 14% caused by family
structure net of all other factors.
Looking at SRP factors,
students are less likely to be deviant if they are female and nonEnglish speakers as well as if they spend more time on homework and
extracurricular activities.
The standardized rank shows that being
female and not being prepared for class have the most impact on
deviance.
At the school level, no significance was found using logistic
regression
except
insignificant
with
variables
the
private
include
class
school
size,
variable.
school
The
problems,
negative school environment, and number of school rules.
The
standardized rank shows that among school level variables, the
private school factor has the most impact on deviance.
For family level factors, odds of being deviant decrease for
those with higher SES percentiles, those who have a high level of
34
parent involvement with the student, and for those whose parents
help them with homework.
Odds of being deviant increase for those
who come from a single-parent family, have a high number of
parental rules, and have siblings who have dropped out. The
variables for urban residence and number of siblings are not
statistically significant.
The standardized rank at the family
level shows that parent involvement has the greatest impact on
deviance.
5. Discussion
It can be seen from the results section of this literature
that several clear conclusions can be drawn when it comes to the
relation between deviance and the three levels of independent
variables used in this study, as seen in the Alternative Model
(2.4).
Those three levels consist of SRP (including sex, minority
status, language, and handicaps), school level factors (including
class size, school problems, school environment, and school rules),
and family level factors (including family structure, socioeconomic
status, parents’ involvement, parent rules, and parent help with
homework).
Using previous studies and the given Alternative Model,
the following hypotheses were generated:
1a. Males are more likely to be deviant than females, net of
other factors.
1b. Net of other factors, students who are held back a grade
are more likely to be deviant.
2a. Net of other factors, students in schools with increased
rules will more likely be deviant.
35
2b. Net of other factors, higher teacher-students ratios
will increase deviance.
3a. Net of other factors, as family SES rises, deviance
decreases.
3b. Net of other factors, students in two-parent families will
have less deviance than students in single-parent families.
Hypothesis 1a is supported by Table 1.
Among those students
who are deviant, 63% are male and 37% are female.
This hypothesis
is also supported by Table 2 when looking at family structure.
Of
students who are deviant in a two-parent household, 65% are male,
and in a single-parent household, 60% are male.
Hypothesis 1b predicts that students who are held back at
least one grade are more likely to be deviant than those students
who are not held back.
According to Table 1, we can see that among
deviant students, there is a higher percent of handicaps than there
is among non-deviant students.
The next hypothesis states that more deviance will be seen in
school with an increased amount of rules.
We can see that for
deviant students the average number of rules is higher than for
students who are not deviant (Table 1).
We know that a larger class size indicates a high studentteacher ratio.
According to Table 1, the average class size is
smaller for a deviant student than for a non-deviant student which
supports Hypothesis 2b by stating that a high student-teacher ratio
will increase the likelihood of deviant behavior.
36
Hypothesis 3a predicts that a higher SES will show a smaller
amount of deviance.
Table 1 supports this prediction with 51.7%
being the average SES percentile of families for students who are
not deviant.
Finally, the last hypothesis is supported in Tables 1, 2, and
3. It is clear in all three analyses that a higher amount of
deviance occurs in single-parent households than in two-parent
households.
Just like any other study, this study has some limitations.
One
limitation
involves
the
operationalization
deviance, which is the dependent variable.
of
the
term
Since the definition of
this variable is provided by the school, it is somewhat biased in
regards to institutional discipline, and it lacks meaning of other
forms of deviant behavior in society.
Another limitation involves
the classification of students who are or are not reported as
deviant.
Those students who are classified as deviant are the ones
who got caught in some act of deviance.
Most likely, there are
many more acts of deviance that are not counted for simply because
the adolescents are not caught doing them.
As seen in this study, adolescents’ educational success is
often
affected
by
deviance.
School
administrators
need
to
implement incentive programs (such as rewarding students for being
prepared for class) to keep students in school and increase their
performance.
Community mentoring programs need to be available for
37
students who receive minimal support from their parents regarding
their education.
Structural level changes in policies need to be
made to make it possible for parents to spend more time with their
children and be more involved in their activities.
Until then, no
improvements will be seen in student role performance.
individuals
begin
to
understand
and
acknowledge
the
When
existing
relationship between student role performance and deviance, then
perhaps future studies will be able to identify an increase of
individual success in our society.
38
BIBLIOGRAPHY
39
Ardelt, Monika & Day, Laurie. 2002. “Parents, Siblings, and
Peers: Close Social Relationships and Adolescent Deviance.”
Journal of Adolescence, 22(3), 310-349.
Bennett-Johnson, Earnestine. 2004. “The Root of School Violence:
Causes and Recommendations for a Plan of Action.” College
Student Journal, June.
Bonikowske, Dennis. 1987. Truancy: A Prelude to Dropping Out.
National Education Service, New York.
Byrd, Robert S., Weitzman, Michael & Auinger, Peggy. 1997.
“Increased Behavior Problems Associated With Delayed School
Entry and Delayed School Progress”. Pediatrics, 100(4),
654-661.
Chyung, Y., Darling, N., & Caldwell, L. 1998. Parental monitoring
and legitimacy of parental authority: flip sides of the same
coin? Biennial Meeting of the Society for Research on
Adolescence, San Diego, CA.
Cook, Bryan G. & Semmel, Melvyn I. 1999. “Peer Acceptance of
Included Students with Disabilities as a Function of
Severity of Disability and Classroom Composition.” Journal
of Special Education, 33(1), 50.
Cresswell, Robert & Rasmussen, Patty. 1996. “Developing a
Structure for Personalization in the High School.” National
Association of Secondary School Principals, NASSP Bulletin,
December.
Crosnoe, Robert, Erickson, Kristan G., & Dornbusch, Sanford M.
2002. “Protective Functions of Family Relationships and
School Factors on the Deviant Behavior of Adolescent Boys
and Girls: Reducing the Impact of Risky Friendships”. Youth
& Society, 33, 515-544.
Dekker, Marielle C. & Koot, Hans M. 2003. “DSM-IV Disorders in
Children with Borderline to Moderate Intellectual
Disability. II: Child and Family Predictors.” Journal of the
American Academy of Child & Adolescent Psychiatry, 42(8),
923-931.
40
Dornbusch, Sanford M., Erickson, Kristan & Laird, Jennifer. 2001.
“The Relation of Family and School Attachment to Adolescent
Deviance in Diverse Groups and Communities”. Journal of
Adolescent Research, 16(4), 396-422.
Downey, Douglas. 1995. “When Bigger Is Not Better: Number of
Parental Resources and Educational Performance.” American
Sociological Review, 60, 746-761.
Downey, Douglas. 1994. “The Educational Performance of Children
in Single-Mother and Single-Father Families: Interpersonal
or Economic Deprivation?” Journal of Family Issues, 15,
129-147.
Downs, William, Robertson, Joan & Harrison, Larry. 1997. “Control
Theory, Labeling Theory, and the Delivery of Services for
Drug Abuse to Adolescents.” Adolescence, Spring.
Epstein, Erwin. 1972. “Social Class, Ethnicity and Academic
Achievement: A Cross-Cultural Approach.” The Journal of
Negro Education, 41(3), 202-215.
Erickson, Kristan, Crosnoe, Robert & Dornbusch, Sanford M. 2000.
“A Social Process Model of Adolescent Deviance: Community,
Social Control, and Differential Association Perspectives”.
Journal of Youth & Adolescence, 29(4), 395-425.
Farrelly, Cathy. 2001.
February 4.
“Public Versus Private School.” The Age,
Garasky, Steven. 1995. “The Effects of Family Structure on
Educational Attainment: Do the Effects Vary By the Age of
the Child?” American Journal of Economics & Sociology,
January.
Garry, Eileen M. 1996. “Truancy: First Step to a Lifetime of
Problems” Juvenile Justice Bulletin, October.
Greenberg, Joel. 1985. “Family Size Tied to SAT/IQ Scores”.
Science News, June.
Hofferth, Sandra, Smith, Julia, McLoyd, Vonnie & Finkeistein,
Jonathan. 2000. “Achievement and Behavior Among Children of
Welfare Recipients, Welfare Leavers, and Low-Income Single
Mothers.” Journal of Social Issues, Winter.
41
Johnson, Monica, Crosnoe, Robert & Elder Jr., Glen. 2001.
“Students’ Attachment and Academic Engagement: The Role of
Race and Ethnicity.” Sociology of Education, 74(4), 318-340.
Juvonen, Jaana. 2001. “School Violence: Prevalence, Fears and
Prevention.” RAND Education: Objective Research and Policy
Analysis, http://www.rand.org/publications/IP/IP219.
Mason, Alex & Windle, Michael. 2002. “Gender, Self-Control, and
Informal Social Control in Adolescence: A Test of Three
Models of the Continuity of Delinquent Behavior”. Youth &
Society, 33, 479-514.
Matherne, Monique & Thomas, Adrian. 2001. “Family Environment
as a Predictor of Adolescent Delinquency.” Journal of
Adolescence, Winter.
Murray, Gloria. 2000. “Class Size: Major Implications for School
Leaders.” National Association of Secondary Principals,
NASSP Bulletin, April.
Pappalardo, William. 2002. “Proofing Schools Against Vandalism–
School Safety.” American Association of School
Administrators, June.
Redmond, Sean M. & Rice, Mabel L. 1998. “The Socioemotional
Behaviors of Children with SLI: Social Adaptation or Social
Deviance?” Journal of Speech, Language & Hearing Research,
41(3), 943.
Riala, K., Isohanni, I., Jokelainen, J., & Jones, P. 2003. “The
Relationship Between Childhood Family Background &
Educational Performance, with Special Reference to Single
Parent Families: A Longitudinal Study.” Social Psychology
of Education, 6(4), 349-365.
Rothstein, Richard. 2000. “The Behavior of Public and Private
Schools–Educational Accountability.” American Association
of School Administrators, May 1.
Sampson, Robert & Raudenbush, Stephen. November 1999. “Systematic
Social Observation of Public Spaces: A New Look at Disorder
in Urban Neighborhoods.” American Journal of Sociology,
105(3), 603-651.
Schaller, Lyle. 1979. “Public Versus Private Schools: A Divisive
Issue for the 1980s.” Christian Century, November, p. 1086.
42
School Violence Resource Center. 2002. National Center for Rural
Law Enforcement–Criminal Justice Institute, University of
Arkansas: Little Rock, AR.
Sewell, William, Shah, Vimah. April 1968. “Parents’ Education
and Children’s Educational Aspirations and Achievements.”
American Sociological Review, 33(2), 191-209.
Shore, Kenneth. 2001. “Success for Students: 12 Practical Tips
To Help Second Language Learners.” Instructor, March.
Sugai, George, Sprague, Jeffrey, Horner, Robert & Walker, Hill.
2000. “Preventing School Violence: The Use of Office
Discipline Referrals to Assess and Monitor School-Wide
Discipline Interventions.” Journal of Emotional and
Behavioral Disorders, Summer.
US Government. “No Child Left Behind Act of 2001”. Education.
Inter-governmental relations, 20 USC 6301 note.
Wells, Frank. 2001. “Zip Codes Shouldn’t Determine Our
Students’ Future.” California Educator, 5(2).
Wenglinsky, Harold. 1997. “How Money Matters: The Effect of
School District Spending on Academic Achievement.”
Sociology of Education, 70, 221-237.
Wright, David. 2005. Sociology 801, Wichita State University.
43
APPENDICES
44
TABLE 1
Values for Full Sample by Deviance
Variables:
Full Sample
Deviant
1
2
***
^
Non-Deviant
Dependent Variable:
%Deviance (0,1) (mean):
40%
(stddev):
0.49
%Female (0,1)
50%
0.50
0.48
%Minority (0,1)
32%
35%
0.47
0.48
%Non-english (0,1)
12%
10%
0.32
0.30
%Held back (0,1)
13%
16%
0.34
0.37
%Remedial coursework (0,1)
11%
13%
0.31
0.34
11%
14%
0.31
0.34
Independent Variables:
SRP factors:
%Handicap (0,1)
Hours per week homework
37%
10.2
9.0
8.98
8.48
Not prepared for class(1-4)
1.88
2.06
0.80
0.76
Hours in extra curricular activity
4.65
4.02
5.78
5.53
%Enrolled in college prep program (0,1)
18%
16%
0.38
0.36
Hours student works per week
11.86
13.93
Hours spent on tv/dvd
14.90
15.96
9.94
10.69
5.19
5.39
School-level factors:
%Private school (0,1)
8%
11%
0.27
0.31
Class size
347.85
336.98
% College prep
200.61
199.98
60%
59.6%
59%
0.49
***
30%
***
13%
***
11%
0.46
0.34
0.32
***
9%
0.29
***
9%
0.29
***
^
11.1
***
^
1.75
9.20
0.80
***
5.07
***
20%
5.91
0.40
***
^
10.47
***
^
9.437
13.98
5.00
***
6%
***
355.13
*
60.9%
0.24
200.71
29.94
29.86
School Problems(1-5)
2.39
2.38
0.34
0.35
0.34
Negative school environment(1-4)
1.73
1.72
1.74
0.57
0.56
0.58
Number of school rules(0-12)
4.67
4.72
4.64
2.21
2.24
2.19
45
29.98
**
2.40
Continued on next page.
46
Family-level factors
%Rural residence (0,1)
21%
21%
0.40
0.41
0.40
%Urban residence (0,1)
28%
29%
28%
0.45
0.45
0.45
%Single Parent Family (0,1)
23%
26%
***
21%
0.42
0.44
Number of siblings
2.32
2.37
**
2.29
1.52
1.54
1.50
Years in neighborhood
10.79
10.69
10.86
8.75
8.96
SES Percentile
50.3%
48.3%
28.87
28.35
3.12
3.08
0.48
0.49
Parents advise student(1-3)
2.23
2.21
0.58
0.60
Parents involved w/ school(0-5)
1.84
1.78
2.29
2.28
Number of parental rules(1-8)
6.89
6.96
1.24
1.19
0.21
0.26
0.64
0.72
Family has resources(0-10)
6.43
6.30
3.15
3.24
Parent helps w/ homework(0,1)
0.46
0.42
0.50
0.49
Days a week student has one meal w/ family
5.34
5.18
1.81
1.85
1.78
Sample n (weighted):
10,886
100%
6,517
59.9%
4,370
40.1%
Parents involved w/ student(1-4)
Number of siblings dropped out
1 = *** p < 0.0001; ** p < 0.01; * p < 0.05
2=effect size greater= >.20
47
20%
0.41
8.61
***
51.7%
29.13
***
3.15
***
2.25
*
1.88
***
6.85
0.47
0.57
2.29
1.28
***
0.18
**
6.51
***
0.49
***
5.44
0.58
3.09
0.50
TABLE 2
Values for Full Sample by Family Structure
Not Deviant
1
Variables:
TwoParent
Independent Variables:
SRP factors:
%Female (0,1)
58%
*
0.49
***
Deviant
2
^
1
SingleParent
TwoParent
61%
35%
0.49
0.48
45%
30%
2
SingleParent
**
40%
0.49
%Minority (0,1)
26%
0.44
0.50
0.46
0.50
%Non-english (0,1)
13%
12%
10%
9%
%Held back (0,1)
10%
0.34
%Remedial coursework (0,1)
%Handicap (0,1)
Hours per week homework
***
0.29
***
^
22%
0.36
0.35
0.41
10%
13%
13%
0.28
0.30
0.34
0.34
9%
9%
12%
0.29
0.28
0.33
10.6
9.2
9.19
8.53
1.9
2.0
11.2
1.72
Hours in extra curricular activity
5.34
**
***
0.78
***
5.95
20%
**
0.40
10.20
**
13.63
Hours spent on tv/dvd
47%
9%
Not prepared for class(1-4)
Hours student works per week
0.30
14%
^
0.30
9.20
%Enrolled in college prep program (0,1)
0.32
15%
***
9.34
**
4.98
Continued on Next Page.
48
^
***
18%
0.38
**
8.3
8.30
*
2.1
0.86
0.76
4.07
4.31
0.78
5.66
5.67
16%
16%
0.37
0.37
0.35
11.48
14.04
13.62
15.19
16.00
9.81
10.54
5.07
5.27
***
^
3.2
5.05
*
14%
15.83
**
11.10
5.67
School-level factors:
%Private school (0,1)
Class size
% College prep
School Problems(1-5)
4%
12%
0.25
7%
0.20
0.33
0.25
353.25
362.12
335.15
342.12
201.59
197.33
200.59
58.5%
61%
29.98
29.89
29.58
2.40
2.42
2.37
0.35
0.33
0.35
**
1.79
1.71
0.59
0.56
**
4.81
4.65
2.27
2.22
2.28
18%
22%
19%
61.7%
Negative school environment(1-4)
1.73
Number of school rules(0-12)
4.59
**
***
0.58
2.16
Family-level factors
%Rural residence (0,1)
21%
%Urban residence (0,1)
25%
Number of siblings
2.26
**
0.41
***
^
0.43
***
1.47
Years in neighborhood
11.11
***
8.31
SES Percentile
54.8%
***
^
***
^
28.76
0.38
0.41
37%
27%
10.31
8.57
40.4%
51.50%
27.67
28.24
3.07
3.11
2.29
2.25
2.30
6.86
6.81
6.99
Number of siblings dropped out
0.16
Family has resources(0-10)
6.75
Parent helps w/ homework(0,1)
0.51
Sample n (weighted):
0.60
0.58
1.57
1.89
1.31
1.17
0.27
0.23
0.69
0.70
5.65
6.53
3.07
3.24
0.42
0.43
0.49
0.50
9.95
***
^
***
^
39.2%
26.59
3.00
0.51
**
2.15
***
1.46
**
6.87
**
0.32
0.65
2.19
1.24
0.76
***
^
5.66
3.16
**
0.38
0.49
5.182
5.29
1.72
1.96
1.80
1.95
5,131
47.1%
1,386
12.7%
3,221
29.6%
1,149
10.6%
1 = *** p < 0.001; ** p < 0.01; * p < 0.05
2=effect size greater= >.20
49
***
33%
10.82
Number of parental rules(1-8)
5.51
0.40
***
9.57
1.97
0.50
0.58
9.94
Parents involved w/ school(0-5)
Days a week student has one meal w/
family
4.89
1.58
0.48
***
**
0.34
1.52
2.23
3.05
1.78
1.60
0.51
^
***
0.47
2.23
***
2.40
2.39
0.46
0.55
30.28
**
0.44
2.25
***
55.6%
2.361
Parents advise student(1-3)
1.27
198.24
***
0.48
3.17
***
7%
2.43
Parents involved w/ student(1-4)
0.56
***
***
^
4.88
TABLE 3
Logistic Regression Analysis for the Deviance Model
Variables:
SRP:
Female (0,1)
Minority (0,1)
Non-english (0,1)
Held back (0,1)
Remedial coursework (0,1)
Handicap (0,1)
Hours per week homework
Not prepared for class(1-4)
Hours in extra curricular activity
Hours student works per week
Hours spent on tv/dvd
(deviant=1)
Full Sample
Two Parent
odds stdz.
odds stdz.
1
ratio rank unstd. 1 ratio rank
unstd.
-0.744
0.168
-0.582
0.007
0.284
0.140
-0.017
0.426
-0.021
0.015
0.024
***
***
***
***
**
***
***
***
***
***
0.48
1.18
0.56
1.01
1.33
1.150
0.98
1.53
0.98
1.02
1.02
Private school (0,1) 0.981 ***
Class size
0
School Problems(1-5) -0.003
egative school environment(1-4) -0.050
Number of school rules(0-12) 0.017
2.67
1
1
0.95
1.02
-0.37
0.08
-0.19
0.09
0.04
-0.15
0.34
-0.12
0.22
0.12
-0.759
0.211
-0.664
-0.068
0.331
-0.025
-0.016
0.463
-0.022
0.017
0.023
2
*** 0.468 -0.38
*** 1.235 0.09
*** 0.515 -0.22
0.935
*** 1.393 0.10
0.976
*** 0.984 -0.14
*** 1.589 0.37 ^
*** 0.978 -0.13
*** 1.017 0.25 ^
*** 1.023 0.12
Single Parent
odds stdz.
1
ratio rank
unstd.
-0.703
0.059
-0.369
0.172
0.165
0.572
-0.021
0.333
-0.018
0.009
0.026
***
***
***
***
*
***
***
0.5
1.06
0.69
1.19
1.18
1.77
0.98
1.4
0.98
1.01
1.03
0.811 ***
0
0.103
-0.041
0.019
2.25
1
1.11
0.96
1.02
*
-0.35
-0.11
0.19
-0.19
0.28
-0.10
0.14
0.14
School:
0.27
1.007 *** 2.738 0.29
0
1
-0.036
0.965 -0.01
-0.047
0.954 -0.03
0.016
1.017 0.03
0.18
Family:
Single Parent Family 0.135 ** 1.14 0.00
0.94
0.002
1.002
-0.180
Urban residence (0,1) -0.060
0.99
0.006
1.006
-0.049
Number of siblings -0.006
SES Percentile -0.002 *
1 -0.06 -0.002 ** 0.998 -0.06
-0.001
-0.306 ***
Parents involved w/ student(1-4) -0.306 *** 0.74 -0.15 -0.312 *** 0.732 -0.15
0.119 *** 1.127 0.15
0.079 *
Number of parental rules(1-8) 0.106 *** 1.11 0.13
0.139 *** 1.149 0.08
0.082
Number of siblings dropped out 0.121 *** 1.13 0.08
-0.062
Parent helps w/ homework(0,1) -0.170 *** 0.84 -0.08 -0.202 *** 0.817 -0.10
(Constant): -0.676 ** 0.51
-0.784
0.457
-0.295
Model Chi-sq 1483.1
542.45
339.41
n= 10,886
8,352
2,534
1=*** p < 0.001; ** p < 0.01; * p < 0.05
2=significant difference between two-parent and single-parent families at the 0.05 level or higher
50
0.84
0.95
1
0.74 -0.16
1.08 0.10
1.09
0.94
0.75